ADS1293EVM: Channel 3 Signal is Noisy in 5 Lead ECG

Hi,

 When you refer to Channel 3, What are the input pins? Do you mean the Input pins IN5 and IN6? Can you tell me what are connected to the input pins for Channel 3?

There might be connection issue.

Did you follow the EVM user guide 2.3.2. 5-Lead ECG Application? How was the IN6 and WCT pin connected?

Thanks

@ChienChun Yang thank you so much for the prompt reply.

ChienChun Yang said:

 When you refer to Channel 3, What are the input pins? Do you mean the Input pins IN5 and IN6? Can you tell me what are connected to the input pins for Channel 3?

Yes the input pins IN5 and IN6. IN5 connected to V1 position and IN6 connected to Wilson's Reference Terminal. The Wilson is computed using IN1, IN2, IN3 and is set by the register values.

ChienChun Yang said:

Did you follow the EVM user guide 2.3.2. 5-Lead ECG Application?

No I totally missed it. But looking at it now, I have some further questions.

Do I have to connect all of the the circled components to the EVM? From the schematics in the datasheet I find out that CVREF already has a 1uF capacitor connected to GND, VDDIO already has 0.1uF capacitor connected to GND and so on. So kindly advise which external components to add.

ChienChun Yang said:

How was the IN6 and WCT pin connected?

By setting WILSONINT to 1

Hi,

May I ask what lead or electrode do you intend to use/read in Channel 3? Precordial leads or Which V#?

And, Are you using a ECG simulator? Does that ECG simulator have function/capability to support the electrode or lead you want to input to Channel 3?
Could you try place/connect to different V# to the input if your ECG simulator support different V#?

Note, some Precordial leads and/or electrodes, depending on the V# and placement, it could be smaller and nosier, so it may require digital filter to filter out noises.

"Do I have to connect all of the the circled components to the EVM?"

Comments: The evm board should come by default as shown in the schematic with those components in the user guide. But, it's good if you could double check they are there. If those components are already there, you don't need to change/modify the EVM.

By referring to datasheet page 20 Figure 20. Wilson Reference Generator Circuit, could you let me know how the SELWILSON1, SELWILSON2, SELWILSON3, WILSONINT and GOLDINT are configured?

The user could program the WILSON_EN1 register to connect the RA electrode to BUF1, program the WILSON_EN2 register to connect the LA electrode to BUF2, and program the WILSON_EN3 register to connect the LL electrode to BUF3.

When the Wilson reference is enabled, its output is present at the WCT pin, except when the analog pace channel is routed to the WCT pin (see Analog Pace Channel). In such a configuration, the Wilson terminal can still be made available at an external pin by programming the WILSONINT bit to 1. Setting this bit connects the output of the Wilson reference internally to the IN6 pin.

Could you try alternatively set WILSONINT bit to 0 and Connect IN6 pin to WCT output externally? And, see if the noises reduced?

ChienChun Yang said:

May I ask what lead or electrode do you intend to use/read in Channel 3? Precordial leads or Which V#?

I am currently using the following cable:

5 Lead ECG Cable Compatible with Schiller Samiks DB9 Clip type – LubdubBazaar

The pins in the DB9 connector of the electrode are in the same order with the connector placed on the ADS1293-EVM. So there is no problem here.

I am planning to use Channel 3 to read the various precordial leads of course one at a time (whichever when deemed necessary by the doctor). At the moment it is connected to V1 output of the cardiac simulator. Here is the link to my cardiac simulator ECG Simulator: Cardiosim III(TI) - Biometric Cables. I am powering it using a 9V Dry Cell.

ChienChun Yang said:

Could you try place/connect to different V# to the input if your ECG simulator support different V#?

I will do this and update here.

ChienChun Yang said:

Note, some Precordial leads and/or electrodes, depending on the V# and placement, it could be smaller and nosier, so it may require digital filter to filter out noises.

Point noted.

ChienChun Yang said:

Comments: The evm board should come by default as shown in the schematic with those components in the user guide. But, it's good if you could double check they are there. If those components are already there, you don't need to change/modify the EVM.

Upon checking I found out that all components are connected. So I am not attaching any external components.

ChienChun Yang said:

By referring to datasheet page 20 Figure 20. Wilson Reference Generator Circuit, could you let me know how the SELWILSON1, SELWILSON2, SELWILSON3, WILSONINT and GOLDINT are configured?

I have left this configuration untouched. It is the default configuration of the 5 Lead ECG which is selected from the dropdown menu in the EVM software. The config is as follows:

SELWILSON1 = 001: IN1

SELWILSON2 = 010: IN2

SELWILSON3 = 011: IN3

WILSONINT = 1: Wilson ref Output = IN6

GOLDINT = 0: Goldberger ref Disable

ChienChun Yang said:

Could you try alternatively set WILSONINT bit to 0 and Connect IN6 pin to WCT output externally? And, see if the noises reduced?

Will try this also and update here.

Hi,

Appreciate for the info.

The following looks right -

"

I have left this configuration untouched. It is the default configuration of the 5 Lead ECG which is selected from the dropdown menu in the EVM software. The config is as follows:

SELWILSON1 = 001: IN1

SELWILSON2 = 010: IN2

SELWILSON3 = 011: IN3

WILSONINT = 1: Wilson ref Output = IN6

GOLDINT = 0: Goldberger ref Disable

"

Do you set the following 

PACE2RLDIN = 0

PACE2WCT = 0

SHDN_PACE = 1

?

Yes, Please let me know how it goes when you connect different V# from the simulator to the IN6.

Also, I saw you write to address   0x11, value = 0x02. May I ask why you want to shut down/set SHDN_CMREF = 1?  Could you also set SHDN_CMREF = 0 and then compare the signal quality difference?

In addition, referring to datasheet page 48, set SHDN_RLD and SELRLD to 0, and try different combination of settings for RLD_BW, RLD_CAPDRIVE to learn if any combination could yield better signal quality.

Do you connect the RL electrode to IN4 pin as in Figure 86? or did you connect RL electrode to RLDOUT pin?

If you connect the RL electrode to IN4 pin, then you need to set SELRLD to 100. Otherwise, try connect RL electrode to RLDOUT pin to observe&compare the signal quality.

I see you post other questions in other tickets. But, let's solve one problem at a time in a step-by-step manner. And, I will reply your questions in other thread.

Thanks

I loaded the default 5 lead config of the ADS1293EVM, then I set the following

All the INA are set to HI_RES mode
The SDM is set to 102400
R1 rate is set to 0 (Standard Pace)
R2 rate is set to 4
R3 rate is set to 32 for all channels

PACE2RLDIN = 0

SHDN_PACE = 1

PACE2WCT = 0

SHDN_CMREF = 0

I connected the the V lead to different positions of my simulator. The first graph corresponds to V4 position and the second graph corresponds to V1 position.

ChienChun Yang said:

Do you connect the RL electrode to IN4 pin as in Figure 86? or did you connect RL electrode to RLDOUT pin?

If you connect the RL electrode to IN4 pin, then you need to set SELRLD to 100. Otherwise, try connect RL electrode to RLDOUT pin to observe&compare the signal quality.

Yes. And SELRLD is set to 100.

ChienChun Yang said:

Also, I saw you write to address   0x11, value = 0x02. May I ask why you want to shut down/set SHDN_CMREF = 1?  Could you also set SHDN_CMREF = 0 and then compare the signal quality difference?

If SHDN_CMREF = 1, even noisier results are obtained. So I will leave it to 0 for all further trials. 

ChienChun Yang said:

In addition, referring to datasheet page 48, set SHDN_RLD and SELRLD to 0, and try different combination of settings for RLD_BW, RLD_CAPDRIVE to learn if any combination could yield better signal quality.

SHDN_RLD is set to 0 for the above two graphs i.e. RLD was enabled and connected to IN4. If I set SELRLD to 0 and disconnect it from IN4, I am obtaining the following

But finally I figured out what the problem was. All this while, my laptop was connected to the Mains power supply via the AC adapter. Just when I realized this and disconnected, I obtained the following graph

This amount of noise is acceptable for my purpose. So my issue is resolved. Thank you for the constant support. Now need to debug the baseline wander issue.